Browsing Versus Searching in Design
Searching and browsing are different. Most people know this, even though we sometimes use the words interchangeably. But as a quick reminder, browsing is more casual, often done to scratch a curiosity itch, whereas searching is more in-depth, done to better understand a topic or to find something. Searching can also be referred to as Analytical search or research. Around 3.2 billion people have access to the internet, so that means more people are searching and browsing every day than ever before, which means it’s even more important to have designs that support browsing and searching.
Let’s look at an example of how searching and browsing for the same topic can provide different information.
Let’s use Google as an example since it is the most used search engine in the world. If a person is just curious about something and wants to satisfy an information itch about college athletes and if they should be paid or not, they could go to Google, type in should college athletes be paid and they will get ~33,800,000 results. The first few results provide a url, sometimes a date, and a small excerpt with information on the pros and cons of paying college athletes, a couple of news articles, and other results with some information, but most likely nothing in-depth. It would provide enough information for a person to decide whether to click on the link and either scan or read the document to find the information they are looking for. A person is also most likely only going to look at the first few results to find the information they are looking for when browsing. It would most likely satisfy a person’s curiosity, and they would be done. However, if a person wants to conduct research and search for more in-depth information, they could go to Google Scholar instead, type the same query (should college athletes be paid), and get ~133,000 results instead of the around 26 million that Google showed. They will see more details about each result such as file type, who wrote it and where it came from, and a small excerpt from the result. Or if they wanted to search in both Google and Google Scholar, they may use Boolean operators such as quotation marks (reducing the results to ~601,000 on Google and only ~108 on Google Scholar) to reduce the number of results to attempt to find results with more value or closer to what the person is attempting to find. The structure of the search results in Google and Google Scholar are different because the information behaviors for the two search engines are different.
If we go based off of this example, we can see that information architecture can be designed to better support browsing and/or searching. Google’s focus is mainly on browsing, however, they incorporate functions that allow for searching (like allowing Boolean operators). Google Scholar’s focus is mainly on searching. So before we dive into how information architecture can be better designed to support browsing and/or searching, we should answer an important question.
Why is this important?
Why should we care about this? Can’t we just add a search function to a website or something and be done? It covers both searching and browsing right? Well, sure you could do that and be done with it without ever giving search a second thought. But think about all of the great website designs we use. Google, Amazon, many library websites, and more. The reason many of us come back to these sites is because they have great search functionalities (and let’s be honest, pretty great designs as well). And if you take a look at the most visited websites in the world, four of the top ten sites are sites where people search and browse for information (Google, Baidu, Wikipedia, and Yahoo). Two more of the top ten sites are for online shopping (Amazon and Tmall).
And of course, we can all think of examples of bad search design and information architecture as well. If you are a researcher, the ACM Digital Library has a ton of great resources, but they have their own special search operators and don’t support Boolean operators, making it a pain to be able to properly search the site. I honestly don’t even use the site for research; I use Google Scholar and other sites to find links to the content in the ACM library. Even though their database is full of great content and resources, many people have a hard time using their site, so they often use something else.
So why is it important to think about search and browse? Because a good search/browse function and design increases site traffic, user satisfaction, and long-term usage. And who wouldn’t want that? Websites aren’t going anywhere any time soon, so they should be created and designed for long-term use. Where else will people go to look for a majority of information?
Good design means good business. And good design is the whole shebang, what is seen visually and what’s underneath it all. Many sites just add a search box and don’t go any further than that to set it up, but that only does so much (and honestly, not much at that). There are so many things that can be done to give people a better browsing or searching experience. The only catch: it can’t be done in just a few minutes. Just like with anything worthwhile, a good browsing or searching function takes some time to create. But the more you do in the beginning, the easier it is and the less that needs to be done later on.
How to support better browsing and searching
So now let’s get to the good stuff. You know that it’s important to incorporate more than just a simple search box, but what exactly do you do? How do you do it? What do you need to keep in mind?
There are many ways to do this. But first and foremost, you need to make an important decision. Is the focus of your website mainly on browsing, searching, or a combination thereof? To make that decision, think about the content you will be providing and what goals your users will have. Once you’ve decided that, you can move on to how to do it. Most of the things we will talk about you have probably experienced, and maybe just haven’t noticed before.
One way to start the process is by thinking about how people browse and how they search. Browsing is more informal and spur-of-the-moment. Searching is more planned and formal, usually with a specific goal in mind. Remembering this can be helpful when designing browse and/or search into your site (and defending why it should be one way or the other!). Most people use a combination of browsing and searching, but here is a good visual of the two.
Search boxes can, of course, support both searching and browsing. People understand that when they want to look for something by typing in a query, they need to look for that little search box and icon. We see search boxes on most websites, but like we said earlier, many of those search boxes only do so much because they aren’t built, just tacked on near the end of the design process (and anything tacked on to something else most likely won’t be very good). How much functionality the search box has though can push a user more towards browsing or searching. A simple search box, of course, is a bit more geared towards browsing. Many simple searches do not have filtering options, although adding this can of course help a person who is browsing (as long as the content is created to support filtering). Simple search boxes often do not support operators, advanced search, or advanced filtering because in browsing, most people don’t want or use these functions. For a search box that is more geared towards searching, it should support Boolean and proximity operators (depending on the content of course), advanced search functionalities (which can be built to use operators without typing them in the query itself), and filtering of results. This may sound like a lot, but when search is designed into the website and not just as an added feature, the whole experience is different.
How the results are displayed from a search or browse query can also vary, just as the Google example earlier showed. Results for browsing tend to be more shallow, with less metadata showing and more about the site itself showing with a short excerpt. They are fairly succinct. Results for searching should be more in-depth, with more metadata showing, information about the source, and an excerpt from the result. This does not mean that the result should take more space or be longer than a browsing result. It’s just that some information is more important when searching than when browsing. Search is often detail driven and will take more time than browsing, so getting a good understanding of something before clicking on it can be very important. Take a look back at the Google and Google Scholar search results. The overall look is the same, but if you spend a minute looking at the results, you can see what I’m talking about. Both are easy to read, but one provides more descriptive information than the other. And one of the main reasons the content of these results is different is because of metadata.
Using labels and metadata can help support both browsing and searching, just in different ways. If a person is on a website and wants to look at coats, they can either type that into the search box or click on the label in the navigation or page content for coats. In searching, labels can be used to help narrow down the search results in the filter section, help people find categories similar to what they are searching, and find other words that may help them in their search endeavors to formulate new queries. Labels are often embedded in metadata, so for browsing it’s not as important to see the labels, but with searching, seeing the labels in each result or page help people get a better grasp of what they are reading or looking at.
Navigation is important for browsing, but not so much for searching. Many people use navigation when browsing a site because they may not have a defined goal. Navigation allows people to become familiar with a site without having to spend a ton of time exploring it. A label can pique a person’s interest and they may continue to browse. Breadcrumbs can help support navigation by allowing people to better understand where they are, and they can help for search as well. If someone is searching for some content and are directed to another website, if they find the information useful, breadcrumbs can provide a lot of information and help people potentially find more relevant information.
What about the people aspect of browsing and searching?
Of course design won’t solve everything (though that would be pretty cool though if it could). Part of the reason that many people can a hard time searching for specific content is because they haven’t learned how to search. It’s not an intuitive thing, but something that we must learn. Many people type in full sentences, sometimes even paragraphs, to try to find some bit of information, and search boxes and websites don’t always work like that. We also don’t always learn about operators like AND, OR, NEAR, or how to use quotation marks properly for searching, and many people actually find these operators to be counter-intuitive. This leads to a lot of frustration that we can only do so much about when designing for browsing and searching. So we should try to educate people on how to search and browse more effectively, and we can also create search functions that take in to account different possibilities of what people may mean and can still effectively produce the kind of results that people are looking for. The internet is still in its early stages, so what’s next and how people will adapt will be something to look forward to.
What’s in store for the future
One thing to remember is that searching and browsing on the internet have evolved quite a bit. Originally, searching was carefully done with thought-out queries, ideas of where to search, and all of this was done by experts because searching required operators, knowledge or how to use computers (before they became more user-friendly), and people often had to pay per minute or per query. Now, none of that holds true, and it most likely won’t hold true for the future either.
The Nielsen Norman Group makes a good point: search alone is not enough. But this doesn’t mean that search isn’t any good. Even though less than 5% of a website’s visitors use search functions on most websites, that doesn’t mean it’s not important. A synergy between browsing and searching is important and will likely be in store for us in the future. People will still sometimes want to browse and sometimes want to search, and with the amount of data on the internet increasing every minute, finding that synergy will separate (and has even today) the sites that we regularly use and those we don’t.
There are a lot of unknown unknowns with the internet, and it takes time for human behavior to change; so we must adapt to people and design search and browse functions for users, not what we want or think will be the “next big thing”. Maybe it’s something like elastic search. Voice search may be the future, maybe digital assistants will also the future of search. Even though typing may not be the primary way of searching and browsing in the future, what’s been discussed here will still hold true: the importance of labeling content with metadata, how we can make more advanced searches, how results are displayed (or read out loud in the future?), and how to design navigation to better support browsing and searching.
The importance of designing good browsing and searching experiences is clear. We will always need to find information in a variety of ways. People will continue to both browse and search, and depending on what a website is about, one or both should be supported. Just as we design and develop websites, their navigation, their designs, and their content, we must also design them for browsing and searching in mind. As our need for information continues to grow, so will our need for better ways to find information also grow.
If you’re interested in learning more about search and browse design, check out the free book Search User Interfaces. It’s a great and very informative book that is free online to read. It inspired this post.